Cycloadditions. XXIV. Cycloadducts from ... - ACS Publications

46-2; 2i, 18887-47-3; 2\-carbonyU*0, ... 18902-55-1; ... Arylaroyldiazomethanes and Carbon Disulfide1,2. John E. ..... Wagner-Meerwein rearrangements ...
8 downloads 0 Views 338KB Size
NOTES

724

The Journal of Organic Chemistry

enrichment was calculated by taking one-half the height of the mass 46 peak times 100 and dividing this result by the sum of the heights of the mass 44 peak and maas 46 peak. The value obtained from the blank sample was subtracted from the value for the enriched sample to compensate for the natural abundance of '*O. Duplicate analyses of the same compounds showed that the analyticd procedure was accurate to &3%. 180

2j, 38887-48-4; 4a, 18893-99-7; 4c, 18894-00-3; 4i, 18902-55-1; Sa, 18894-01-4; Sc, 18894-02-5; Si, 18894-03-6,

Acknowledgments.-The authors are most grateful to Mr. Gary Weber for performing the mass spectral analyses, to Professor Marvin Charton for several Registry No.--2a, 18888-64-7 ; 2a-~arbonyL'~O, helpful discussions, to the Arizona State University 18893-96-4; 2b, 18887-40-6; 2 ~ 9 16544-65-3; 2 ~ - Grants Committee for financial support and two summer stipends for K. B. G., and to the Garrett Corp. carbonyl-180, 18893-97-5 ; 2d, 18887-42-8; 2e, 18887and Arizona State University for use of their computer 43-9 ; 2f, 18887-44-0; 2g, 18887-43-9; 2h, 18887facilities. 46-2; 2i, 18887-47-3; 2i-c~rbonyL~~0,18893-98-6;

Notes is distinct. This factor became the basis for an approach to the structural distinction; the adduct in which R = p-CeH4F- (2c or 3c) was synthesized and its fluorine nmr spectrum examined. JOHN E. jBALDWINa AND JONA. KAPECKI' p-Fluorophenyl-p-fluorobenzoyldiazomethane (IC) was synthesized through the sequence p-fluorobenxDepartment of Chemistry a d Chemical Engineering, University o j Illinois, Urbana, Illinois 61801 aldehyde + p,p'-difluorobenzoin + p,p'-difluorobenzil -t p,p '-difluorobenzil monohydrazone + IC. Received July 9,1968 Heating the difluoroazibenxil and carbon disulfide to reflux gave a mixture of products from which the adduct A reinvestigation of the adduct derived from aziCzsHleF40zSz 2c or 3c, mp 145-146", was isolated by benzil (la) and carbon disulfide6led Yates and Christenfractional crystallization. sen6 to assign it structure 2a (or its geometrical isomer). The infrared spectra of this tetrafluoro adduct, mp A supposedly analogous adduct obtained from the re145-146' , and of the adduct from azibenzil were identiaction of 3-diazobutanone (lb) and carbon disulfide cal but for three exceptions. One was the expected'O was represented by formula 2b,' but a recent X-ray strong carbon-fluorine bond absorption a t 1162 cm-l. single-crystal structure analysis on this adduct estabThe second was a weak band at 945 cm-' in the tetralished 3b as an accurate structural a s ~ i g n m e n t . ~ ~fluoro ~ adduct, not observed in its unsubstituted analog. The third difference came in the region 1620-1575 cm-l. The compound assigned structure 2a shows three medium-intensity peaks a t 1620, 1598, and 1578 RC-CR cm-'. In the tetrafluoro adduct, this region is obscured by a very strong band at 1598 cm-' with a 1 2 3 shoulder probably containing a band near 1620 cm-'. The adduct of mp 145-146' showed three absorpa, R = CdH:, tions in a l9F nmr spectrum; they had similar fine b,R=CHI structure and a 1 : 1 :2 ratio of relative intensities (Figure c , K = pC,,H,F 1). In tetrahydrofuran, these absorptions appeared Whether the adduct from azilbenzil might require a similar structural reassignment (3a instead of 2a) was answered through the present work.

Cycloadditions. XXIV. Cycloadducts from Arylaroyldiazomethanes and Carbon Disulfide's2

P

. I

Results and Discussion Structure 2 has two R groups which would be identical by symmetry, whereas in 3 each of the four R groups (1) Supported in part by Public Health Reaearch Grant GM-14381. (2) Paper XXIII in this series: J. E. Baldwin and J. E. Gano, J . 0rp. Chcm., in press. (3) Alfred P . $loan Research Fellow. (4) National Science Foundation Predoctoral Trainee, 1985-1968. (5) J. Meyer, Helv. Ch,im. Acta, 0 , 38 (1925). (6) P. Yatea and B. G. Christensen, Chem. Znd. (London), 1441 (1958). (7) A. J. Kirby, Tetrahedron, 99, 3001 (1968). (8)J. A. Kapecki, J. E:. Baldwin, and I. C. Paul, Tetrahedron Lett., 5307 (1987). (9) J. A. Kapecki, J. E. Baldwin, and I. C. Paul, J . Amer. Chem. Soc., $0, 5800 (1988).

101

'9F

I

D)

Spectrum

941 MHZ

Figure l.-Fluorine-19 nmr spectrum of adduct 2c, mp 145-146', taken as a solution in tetrahydrofuran. (10) K. Nakanishi, "Infrared Absorption Spectroscopy," Holden-Day, Inc., San Francisco, Calif., 1982, p 67.

NOTES 725

Vol. 34, No. 3, March 1969 at 106.4, 111.8, and 113.1 ppm relative to internal fluorotrichloromethane. In bromoform, these peaks came a t 105.3, 111.0, and 112.3 ppm, and no broadening of the upfield signal was observed. Fluorine chemical shifts are particularly sensitive to solvent effects;11*12 the constancy of the 1:1:2 pattern as solvent and chemical shifts change affords some indication that an accidental coincidence of fluorine chemical shifts in two nonidentical p-C6H&' groups is not responsible for the spectral resulk The spectroscopic data for the tetrafluoro adduct are consistent with structure 2c, not with 3c, and the adduct from azibenzil is accordingly better represented by 2a than by 3a. The arylaroyldiazomethanes and carbon disulfide give adducts of structure 2, while the alkylalkoyldiazomethanes lead to the 2-methylenel,&dithiacyclobutanes 3. Fresh degradative and synthetic results reported by Yates and Williams13 have independently corroborated structure 2a for the adduct from azibenzil.

very similar to that of adduct 2c, except for the absence of a strong band a t 1710 cm-'. The spectra of A and B were virtually identical except for minor variations : a t 1560 and 987 cm-' the high melting isomer B has bands of medium intensity not distinctly present in the other; the relative intensities of doublets at 1095 and 1080 cm-' differ in the two adducts; a band a t 900 cm-l appears in the spectrum for B, while A shows a corresponding band a t 885 cm-'. Isomer B had X , C ~ ~ c l 267 z mp (e 3.2 X lo3) and 391 (4.7 X lo3). Low-resolution mass spectra of the two adducts taken under similar conditions show nearly identical fragmentation patterns (Figure 2). High-resolution mass spectra for adduct A,14 summarized in Table I, permitted sure assignments of empirical formulas for some peaks. It proved impossible to obtain elemental composition data for the molecular ion at m/e 580. The m/e 581 peak distorted the P F K peak at 581 so that it could not be used as a mass standard line. Structure postulates 4 and 5 for these adducts were

I / I

I I

.

I

I

1

Figure 2.-Low-resolution mass spectra of CSOH~BF~OZS~ adduch. The isomer of mp 274-275' gave the upper spectrum.

Two other adducts (A, mp 235-236", and B, mp 274-275') were obtained from the reaction of difluoroazibenzil with carbon disulfide. Both had molecular ions in their mass spectra a t m/e 580 and elemental analyses most consistent with the molecular formula CwHieF&% The infrared spectra of these adducts in KBr were

evolved through consideration of the spectral data and conceivable mechanistic routes. These postulates may be used to account for the mass spectral fragmentation data as outlined in Scheme I and may be envisioned as originating through Diels-Alder reactions of 6 and 7.

TABLE I HIGH-RESOLUTION MASSSPECTRAL DATAFOR MAJORPEAKS IN SPECTRUM OF ADDUCT WITH MP 236' Molecular formula

Obsd

Calcd

Re1 intensity

CaoHieFIOzSz CisHsFzOS Ci4H,FzS Ci&H92 CiaH4'z C7HdFO C7HIF CBH~

548.0523 274.0254 246.0323 214.0592 201.0502 123.0272 107.0304 75.0286

548.0528 274.0264 246.0315 214.0594 201.0516 123.0246 107.0297 75.0235

12.3 14.2 6.8 6.8 8.0 38.2 15.1 19.7

m/e

(11) L. M. Jackman, "Applications of Nuclear Magnetic Resonance Spectroscopy in Organic Chemistry," The Macmillan Co., New York, N. Y., 1959, p 76. (12) J. W. Emsley, J. Feeney, and L. H. Sutcliffe, "High Resolution Nuclear Magnetic Resonance Spectroscopy," Pergamon Press Ltd., London, 1966, p 872. (13) P. Yates and L. L. Williams, Tetrahedron Lett., 1205 (1968).

Ar

rlr 5

4

XS Ar

6

7

Ar = p-C6H4F

Conclusions Two clear results were obtained through this investigation. Structure 2 is preferable to structure 3 for the type of cycloadduct obtained by Meyer6 from (14) Obtained through the good offices of Dr. George Van Lear and the Mass Spectrometry Center at Purdue University.

726

NOTES

The Journal of Organic Chemistry

drate in ethanol.” An 81% yield of fine yellow crystals, mp 183185’, was obtained. The analytical sample, mp 190-191”, was recrystallized from ethanol. Anal. Cdcd for CI~HWFZNZO: C, 64.63; H, 3.87; N, 10.77. SCHEME 1 Found: C, 64.33; H, 4.05; N, 10.84. p-Fluorophenyl-p-fluorobenzoyldiazomethane was prepared RATIONALIZATION OF MASSSPECTRUM OF CsaH~@402Ss ADDUCTS from 11.1 g of p,p’-difluorobenzil monohydrazone through oxidation by mecuric oxide.18 The crude product (6.8 g) was -HF recrystallized from diethyl ether to give 3.2 g of orange product: ArC’O Art C6H: mp 102-106’ dec; 8 (CHCls) 2080, 1620-1600, 1500, 1340, m/e 123 m/e 95 m/e75 1155, 865, and 830 cm-1. Further recrystallizations did not raise the decomposition temperature. Anal. Calcd for C I ~ H ~ F ~ NC, ~ O65.13; : H, 3.12; N, 10.84. Found: C, 65.23; H, 3.36; N, 10.81. Reaction of p-Fluorophenyl-p-fluorobenzoyldiazomethane with Carbon Disulfide.-p-Fluorophenyl-p-fluorobenzoyldiazomethane (4.4 g, 17 mmol) and 40 ml of carbon disulfide were heated at reflux for 40 hr. The reaction mixture was cooled, filtered, and m/e 274 concentrated. Yellow crystals (4.4 g) were collected, washed with ether, and dried. The crude product was recrystallized from absolute ethanol to give two crops of yellow crystals, 1.2 and 1.0 g, having mp 90105’, and a residue insoluble in ethanol (1.2 g of yellow material, mp 267-270’). The first two crops of crystals were dissolved in hot ethanol and, as the solution was allowed to cool slowly, a flocculent mass of very fine, hairlike yellow needles was obtained. This solid had mp 235-236’ after repeated recrystallization from ethanol. Anal. Calcd for C30H16F402S3: C, 62.06; H, 2.78. Found: m/e 580 m/e 548 m/e 456 C, 61.43; H, 2.92. -C,H, Concentration of the ethanolic mother liquors gave another Ad!% CjHZFS solid which, after repeated recrystallization from carbon diI m/e 139 m/e113 sulfide, had mp 145-146’. Anal. Calcd for C29HlsO~S2: C, 64.91; H, 3.01; S, 11.95. Found: C, 64.54, 64.58; H, 3.12, 3.15; S, 11.81. The high-melting residue from the first recrystallization was A recrystallized four times from chloroform to obtain lemon yellow places, mp 274-275’ dec. Jo)-St ---t ArCrPrCAr.’ C,H,+F Anal. Calcd for CIIOHIRF‘O~SI: __ -- . . . C, 62.06: H, 2.78; S, 16.58. Ar \ m/e 214 m/e 107 Found: C, 61.55; H, 2.98; S, 16.42. The combined 3.2 g of crops 1, 2, and 3 contained approxim/e 580 mately 12% of the adduct of mp 235-236’, 66% of the CWHMFC 082 adduct, mp 145-146’, and 22% by weight of the high meltproduct, mp 274-275’. Spectral data for these three adducts are given in the discussion. ‘& Ar,CCSf ArC7H3F’ Registry No.-p,p-Difluorobenzil monohydrazone, m/e 246 m/e 201 18598-41-9; IC, 18542-79-5; ZC, 18542-80-8; 4, 18542Ar = pC6H,F 81-9; 5,18542-82-0; carbon disulfide, 75-15-0. Y,Z =OandS

azibenzil and carbon disulfide. Seemingly modest variations in substrate 1, replacing phenyl with methyl,

--

-

1

-

\

\

a

or phenyl with p-fluorophenyl, lead to strikingly different results in reactions with carbon disulfide. The alkylalkoyldiazomethanes give rise to adducts 3 ; p fluorophenyl-p-fluorobenzoyldiazomethane gives both the expected adduct 2c and two more complex products C80H~aF40zS8.Structural determinations for adducts A and B, and an understanding of how they are formed, will depend upon further experimentation. Experimental Section p,p’-Difluorobenzil was prepared in two steps from p-fluorobenzaldehyde, according to a procedure described for the conversion of benzaldehyde to benzoin to benzil.16 p,p’-Difluorobenzoin was recrystallized from ethanol and had mp 85-86’. Anal. Calcd for ClrHloOIFa: C, 67.74; H I 4.06. Found: C, 67.46; H, 4.31. p,p’-Difluorobenzil was recrystallized from benzene to give yellow needles, mp 119-121’ (lit.” mp 123-123.5’), in 38% over-all yield. p,p‘-Difluorobenzil Monohydrazone.--p,p’-Difluorobenzil (13 g) was converted into the monohydrazone with hydrazine hy(15) L. F. Fimer, “Experiments in Organic Chemistry,” 3rd ed, D. C. Heath and Co., Boston. Mass., 1957, p 170. (18) W. Voegtli snd P. Llluger, Hclu. Chim. Acta, $8, 48 (1955).

(17) Cf.L. I. Smith and H. H. Hoehn, “Organic Syntheses,” Coll. Vol. 111, John Wiley & Sons. Inc., New York, N. Y., 1955, p 356. (18) Cf.C. D. Nenitzescu and E. Solomonica, “Organic Synthesas,” Coll. Vol. 11, John Wiley & Sons, Inc., New York, N. Y.. 1943, p 496.

Lack of Oxygen-18 Scrambling in the Solvolytic Rearrangement of Bicyclo[2.2.0]hexane-l-methyl p-Nitrobenzoate18 WILLIAMG. DAUBENAND JAMES L. CHIT WOOD'^

Department of Chemistry, University of California at Berkeley, Berkeley, California 947’30 Received August MI 1968

Wagner-Meerwein rearrangements can be considered to fall within two theoretical limits with regard (1) (a) This investigation wag supported in part by Grant No. GP 3880, National Science Foundation. (b) National Institutes of Health Predoctoral Fellow, 1966-1968.